223. NON-CONTACT ULTRASONIC SYSTEM FOR RAIL FLAW DETECTION AND SURFACE CHARACTERIZATION
Name: Stefano Mariani
Grad Year: 2015
Robert Phillips, email@example.com
As rail freight companies continue to increase the tonnage moved on their rail system, the cost of combating the effects of rolling contact fatigue (RCF) also grows. RCF is the major cause to the formation of Head Checks, Shelling and Flaking on the surface of the rail. If these surface conditions are left unattended, they will become internal transverse defect, a major cause of rail breaks. The most effective way to prevent rail breaks caused by this kind of defects remains a preventive maintenance routine of rail grinding to remove the degraded surface. To assess the severity of these surface and sub-surface defects several methods have been proposed in order to determine the amount of grinding needed to rejuvenate the rail profile. However, the methods being used so far have proven to be not reliable enough. Therefore, the goal of this work is to pursue the development of an alternative technique able to investigate and classify the superficial condition of rails in a non-contact manner expanding of the current system used to detect internal defects. The core of this method characterizes the rail surface cracks by measuring the attenuation (damping) of ultrasonic surface (Rayleigh) waves travelling on the railhead surface. This attenuation is measured at different frequency bands of the waves to probe different depths, and hence providing crack information along the depth of the railhead surface. The frequency-dependent damping of the ultrasonic surface waves can be measured either deterministically or through a statistical Outlier Analysis or Anomaly Detection algorithm that may enhance the classification of crack density and depth. Once the method for surface characterization is validated and integrated into the internal defect system, field tests will be performed to compare the new method to current industry standards.